U.S. patent application number 10/075507 was filed with the patent office on 2002-10-17 for tunable power amplifier matching circuit.
Invention is credited to Forrester, Tim, Toncich, Stanley S..
Application Number | 20020149428 10/075507 |
Document ID | / |
Family ID | 36806192 |
Filed Date | 2002-10-17 |
United States Patent
Application |
20020149428 |
Kind Code |
A1 |
Toncich, Stanley S. ; et
al. |
October 17, 2002 |
Tunable power amplifier matching circuit
Abstract
A power amplifier matching circuit is provided. The matching
circuit includes a ferro-electric tunable component. A control
signal is applied to the tunable component, changing the
component's impedance. This changes the impedance of the matching
circuit.
Inventors: |
Toncich, Stanley S.; (San
Diego, CA) ; Forrester, Tim; (San Diego, CA) |
Correspondence
Address: |
Kyocera Wireless Corp.
Attn: Patent Department
PO Box 928289
San Diego
CA
92192-8289
US
|
Family ID: |
36806192 |
Appl. No.: |
10/075507 |
Filed: |
February 12, 2002 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60283093 |
Apr 11, 2001 |
|
|
|
Current U.S.
Class: |
330/302 |
Current CPC
Class: |
H01G 7/06 20130101; H03B
5/1262 20130101; H03F 2200/111 20130101; H04B 1/40 20130101; H03B
5/04 20130101; H01P 1/20336 20130101; H01L 27/0808 20130101; H01P
1/20363 20130101; H03L 7/18 20130101; H01P 5/04 20130101; H03H 7/20
20130101; H04B 1/0458 20130101; H04B 1/30 20130101; H01P 1/213
20130101; H03H 7/12 20130101; H01P 1/20381 20130101; H01P 1/2056
20130101; H03L 1/022 20130101; H03B 5/124 20130101; H04B 1/0053
20130101; H03H 7/0123 20130101; H03B 5/1841 20130101; H03B 5/362
20130101; G01R 27/2694 20130101; H01P 1/203 20130101; H03F 2200/372
20130101; H03J 5/246 20130101; H03B 5/1293 20130101; H03F 3/191
20130101; H03F 1/56 20130101; H03F 2200/294 20130101; H01P 1/2039
20130101 |
Class at
Publication: |
330/302 |
International
Class: |
H03F 003/191 |
Claims
What is claimed is:
1. A tunable power amplifier comprising: a power amplifier; a
ferro-electric tunable component coupled to the power amplifier; a
power amplifier output matching circuit coupled to the power
amplifier, having an impedance and comprising the ferro-electric
tunable component; a control line operably coupled to the
ferro-electric component; a control source electrically coupled to
the control line, the control source configured to transmit a
control signal on the control line; wherein the ferro-electric
component, responsive to the control signal, adjusts the impedance
of the matching circuit.
2. The tunable power amplifier of claim 1, wherein the
ferro-electric tunable component comprises a ferro-electric tunable
capacitor.
3. The tunable power amplifier of claim 2, further comprising a
substrate wherein the capacitor is directly mechanically coupled to
the substrate and the power amplifier is directly mechanically
coupled to the substrate.
4. The tunable power amplifier of claim 3, wherein the output
matching circuit further comprises a second ferro-electric tunable
component.
5. The tunable power amplifier of claim 4, wherein the second
component comprises a tunable ferro-electric capacitor.
6. The tunable power amplifier of claim 1, wherein the matching
circuit comprises: a first tunable ferro-electric capacitor coupled
at a first end of the first capacitor to an output of the power
amplifier and to ground at a second end of the first capacitor; an
inductive element coupled at a first end of the inductor to the
first tunable capacitor and to the power amplifier, and; a second
tunable ferro-electric capacitor coupled, at a first end of the
second capacitor to a second end of the inductive element and to
ground at a second end of the second capacitor; wherein, the
ferro-electric component comprises one of the ferro-electric
tunable capacitors.
7. The tunable power amplifier of claim 6, wherein the inductive
element comprises a lumped element inductor.
8. The tunable power amplifier of claim 6, wherein the inductive
element comprises a microstrip.
9. The tunable power amplifier of claim 6, further comprising: a
second inductive element coupled at a first end of the second
inductive element to the second end of the first inductive element;
a third ferro-electric tunable capacitor coupled at a first end of
the third capacitor to a second end of the second inductive element
and at a second end of the third capacitor to ground.
10. The tunable power amplifier of claim 9, wherein the second
inductive element comprises a lumped element inductor.
11. The tunable power amplifier of claim 9, wherein the second
inductive element comprises a microstrip.
12. A method of tuning an impedance match of a power amplifier
comprising: generating a control signal; coupling the control
signal to a ferro-electric component; changing an impedance of the
component, responsive to the control signal; changing the impedance
match of the power amplifier responsive to changing the impedance
of the component.
13. A wireless communication device comprising: a battery; a
transceiver; a user interface; a housing encasing the battery and
the transceiver and adapted to present the user interface external
to the housing; a power amplifier; a ferro-electric tunable
component coupled to the power amplifier; a power amplifier output
matching circuit coupled to the power amplifier, having an
impedance and comprising the ferro-electric tunable component; a
control signal generator for generating a control signal; a control
line coupled to the control signal generator and to the
ferro-electric component; a control source electrically coupled to
the control line, the control source configured to transmit a
control signal on the control line; wherein the ferro-electric
component, responsive to the control signal, adjusts the impedance
of the matching circuit.
Description
RELATED APPLICATIONS
[0001] This application claims the benefit of U.S. Provisional
Application No. 60/283,093, filed Apr. 11, 2001, which is hereby
incorporated by reference. In addition, this application relates to
U.S. application Ser. No. 09/904,631 filed on Jul. 13, 2001, Ser.
No. 09/912,753 filed on Jul. 24, 2001, Ser. No. 09/927,732 filed on
Aug. 8, 2001, Ser. No. 09/927,136 filed on Aug. 10, 2001, "Tunable
Capacitor," filed on Jan. 11, 2002, and "Antenna Interface Unit,"
filed on Feb. 8, 2002, which are hereby incorporated by
reference.
BACKGROUND
[0002] Wireless communication devices, such as, but not limited to,
wireless telephones, use many electronic components to transmit and
receive signals over the air. A transceiver is the part of a
wireless telephone that actually sends and receives signals. The
front end of a transceiver is the portion of a transceiver closest
to the air interface in the signal path. The front end includes an
antenna and several components near the antenna in the signal path.
Several of the components required in the front end of the
transceiver are power amplifiers (PA's), isolators, low noise
amplifiers (LNA's) and multiplexers. Each of these components are
typically manufactured as packaged devices. In the case of a PA or
an LNA, this package typically includes the active device and
internal input and output matching circuits for bringing the input
and output resistances up to an industry standard 50 ohms.
[0003] In one common embodiment, the packaged PA is comprised of a
high performance FET (e.g., GaAs) placed on a ceramic or other
substrate. Other active devices can be used, such as, for example,
bipolar junction transistors (BJT's) and high electron mobility
transistors (HEMT's). The matching circuits may be patterned on the
ceramic substrate, or they may be fabricated using lumped surface
mount technology (SMT) components. The FET is bonded to the package
substrate, possibly to a metal heat sink, then typically connected
to its input, output and bias pads using bond wires.
[0004] Depending on the requirements, multi-stage PA devices may be
used as well. This means that one PA device may include more than
one amplifying transistor. This may be necessary for a number of
reasons. One possible reason is to produce the required gain. In
the case of a multi-stage PA device, inter-stage impedance matching
circuits may be used as well, to match between the output of one
stage and the input of the following stage.
[0005] The inputs, outputs and bias lines to the FET are routed
down to the ceramic substrate. After passing through the matching
circuits, the input and output lines are routed off of the
substrate down to the underlying printed wire board (made of FR-4
in most cases) through connectors on the PA package. Further wire
bonding may be required to connect the package pads to the input,
output and bias lines.
[0006] The package further comprises some kind of packaging
(typically polymer) encasing, in whole or in part, the FET and the
ceramic substrate holding the matching circuits. The input and
output bias leads can be found at the edge of the packaging.
[0007] Isolators, duplexers, diplexers and low noise amplifiers
(LNA's) are handled in much the same way. As packaged devices, they
each have their separate substrates with their separate matching
circuits bringing their input and output impedances to 50 ohms.
[0008] Most RF test equipment can only test parts at an impedance
of about 50 ohms. Manufacturers and designers typically want to be
able to test each part separately. Historically, the only way this
could be done was if each part had input and output impedances of
around 50 ohms. For this reason, parts, such as PA's and LNA's, for
example, have typically been manufactured with impedances equal to
about 50 ohms. This has required the use of extensive input and
output matching circuits for many of these parts.
[0009] A duplexer is one of the primary components in a transceiver
front end. The duplexer has three ports (a port is an input or an
output). One port is coupled to an antenna. A second port is
coupled to the transmit signal path of the transceiver. The
duplexer couples the transmit path to the antenna, so that the
transmit signal can be transmitted on the antenna.
[0010] A third port is coupled to the receive path of the
transceiver. The antenna coupled the antenna to the receive path,
so that the received signal can be received by the receive path of
the transceiver.
[0011] An important function of the duplexer is to isolate the
transmit signal from the receive path of the transceiver. The
transmit signal is typically much stronger than the receive signal.
Some of the transmit signal inherently gets down the receive path.
But this transmit signal going down the receive path must be
greatly reduced (or attenuated). Otherwise, the transmit signal
going down the receive path will swamp, or overwhelm, the receive
signal. Then the wireless telephone will not be able to identify
and decode the receive signal for the user.
[0012] The required attenuation of the transmit signal going down
the receive path is achieved at some expense. The duplexer also
attenuates the transmit signal going to the antenna for
transmission. This attenuation in the transmit signal going to the
antenna is known as loss. It would be beneficial to reduce the
transmit path loss in the duplexer.
[0013] Additionally, the duplexer typically must be large
accomplish the receive path attenuation of the transmit signal.
Consumers are continually demanding smaller and smaller wireless
telephones with more and more features and better performance.
Thus, it would be beneficial to reduce the size of the duplexer
while maintaining or improving the transmit signal attenuation in
the receive path and simultaneously maintaining or improving the
transmit signal loss to the antenna.
SUMMARY
[0014] Transceivers account for a significant portion of the cost,
size and power consumption of wireless communication devices. The
front end, including antennas, duplexers, diplexers, isolators,
PA's, LNA's and their matching circuits accounts for a significant
portion of the cost, size and power consumption of the transceiver.
It would be beneficial to reduce the cost, size and power
consumption of these parts, individually and together.
[0015] Briefly, the present invention provides a ferro-electric
tunable duplexer integrated with one or more of the other parts.
This combination is referred to herein as an antenna interface
unit. More specifically, in addition to adding F-E tunability, the
present invention integrates one or more of the above components on
one substrate. The components are integrated on one substrate
either by placing each component, with the appropriate matching
circuit directly on the substrate, or by direct fabrication of the
component and matching circuit into or onto the substrate.
[0016] For example, in the case of integrating the PA, the isolator
and the duplexer, the PA active device (e.g., GaAs FET) is placed
directly onto the common substrate. As part of the integration of
components, the matching circuits for the components may be
patterned or placed on the common substrate. The matching circuits
for the PA would be patterned or placed on this substrate. The
isolator, if used, could be fabricated directly on this common
substrate or mounted as a discrete component.
[0017] The matching circuit between the isolator and the duplexer
would be patterned or fabricated on the substrate. The isolator
would have its junction patterned on this substrate, with the
ferrite puck, magnet and shield placed over it.
[0018] For purposes of integration, a stripline duplexer may be
preferred as it would use the common substrate as one half of each
resonator. Additionally, its length is shorter than a corresponding
microstrip realization. Whatever type of duplexer is used, any
coupling and tuning capacitors would be patterned on the common
substrate. It will be understood that the same kind of integration
can be carried out for the LNA, diplexer and antenna matching
circuits. If minimum loss is a key requirement for a post PA BPF,
duplexer or multiplexer, then a low loss substrate must be used as
is well known to those skilled in the art.
[0019] The topology of the matching circuits would be typical
matching circuit topologies with two key exceptions: (1) they would
be integrated with the other parts and matching circuits on the
common substrate and (2) they may comprise F-E tunable components,
though they need not all comprise F-E tunable components. The PA
and isolator matching circuits would typically be pi matching
circuits (shunt capacitor, series inductor or microstrip line,
shunt capacitor). The isolator typically uses series or shunt
reactive circuits. The diplexer and duplexer matching circuits
would typically be simply series input and output capacitors. The
antenna matching circuit would be a pi or T circuit with L-C
ladders creating a higher order matching circuit. Preferably, the
duplexer would be as claimed in U.S. patent application Ser. No.
09/912,753 filed on Jul. 24, 2001.
BRIEF DESCRIPTION OF THE DRAWINGS
[0020] FIG. 1 is a block diagram of a side view of an antenna
interface unit.
[0021] FIG. 2 is a schematic diagram of a power amplifier matching
circuit.
[0022] FIG. 3 is a schematic diagram of an extended power amplifier
matching circuit.
[0023] FIG. 4A is a schematic diagram of a multiband power
amplifier matching circuit.
[0024] FIG. 4B is a schematic diagram of another multiband power
amplifier matching circuit.
[0025] FIG. 5 is a schematic diagram of an isolator and its three
matching circuits.
[0026] FIG. 6 is a schematic diagram of a LNA matching circuit.
[0027] FIG. 6B is a graph of a LNA noise figure response.
[0028] FIG. 7 is a schematic diagram of an antenna matching
circuit.
[0029] FIG. 8 is a block diagram of an antenna interface unit.
[0030] FIG. 9 is a block diagram of an antenna interface unit.
[0031] FIG. 10 is a block diagram of an antenna interface unit.
[0032] FIG. 11 is a block diagram of an antenna interface unit.
[0033] FIG. 12 is a block diagram of an antenna interface unit.
[0034] FIG. 13 is a block diagram of an antenna interface unit.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
[0035] Referring now to FIG. 1, an integrated antenna interface
unit (AIU) 12 is shown. While a duplexer is shown, as is common in
a CDMA handset, a multiplexer could be used as well. While the
description that follows specifies a duplexer throughout, it is to
be understood that a multiplexer or BPF could be substituted for
the duplexer. A PA unit 20, an isolator unit 24, and a duplexer 28
are all attached to a common substrate 16, eliminating the need for
individual substrates for each of these components.
[0036] The substrate is preferably made of a carefully selected
material. The substrate parameters that are typically critical are
dielectric constant, loss tangent, thermal properties, cost and
ease of processing. Typically, a dielectric constant should be less
than about 40, and the loss tangent should be less than about 0.001
in the frequency range of interest. A low loss substrate may be
more expensive than a higher loss substrate. A designer must
frequently balance the issues of cost and performance parameters
such as loss. Additionally, metal loss must also be minimized. A
substrate must be chosen that can accommodate a low loss metal.
[0037] Advantages of integration of components with a multiplexer
include: (1) reduction of the overall loss associated with the
integrated device compared to that arising from using discrete
parts, thus making it easier to meet specifications; (2) reduction
of the footprint of the Tx chain in the sybsystem; (3) reduction of
the overall parts count, especially as far as a manufacturer of a
wireless communication device is concerned; (4) reduction of cost
due to reduced packaging and parts count, (5) integration of f-e
tunable components with lower added loss and occupying less space
than if introduced as individual, lumped element components.
[0038] A PA-to-isolator matching circuit 41, disposed on the
substrate 16, couples the PA unit to the isolator unit 24. An
isolator-to-duplexer matching circuit 44, disposed on the substrate
16, couples the isolator to the duplexer.
[0039] Preferably, an isolator is used, but it is optional. If no
isolator is used, it will be understood that the isolator is
removed and the PA-to-isolator matching circuit and the
isolator-to-duplexer matching circuit is replaced by a
PA-to-duplexer matching circuit. There are two main reasons a
designer will choose to use an isolator in a design such as
disclosed herein. The reasons are: (1) To provide a certain load
impedance to the device preceding the isolator (the PA in this
case); and (2) To prevent unwanted signals from propagating back
into the device preceding the isolator (the PA in this case).
Unwanted signals propagating back to the PA can cause unacceptable
mixing or distortion or both to be created which can render the
overall design unacceptable.
[0040] As is well known in the art, there are many cases where the
isolator can be eliminated. This is true when: (1) The PA can be
presented an acceptable load under operating conditions; or (2) The
isolator can be replaced by a suitable coupler or passive hybrid
device that reduces the effect of reverse power propagation on the
desired signal path. Passive couplers or passive hybrid couplers
can be more easily implemented by direct fabrication on the
substrate as outlined in this application.
[0041] This particular configuration of the AIU 12 is shown and
described in detail for purposes of example and illustration only.
The AIU 12 may not include a PA 20 or an isolator 24. As will be
described more generally with reference to FIGS. 8-13, the AIU
always has a tunable multiplexer and some other component on a
common substrate. Other than that, there are many possible
components that can be integrated with the multiplexer to form the
AIU. The PA and isolator are just two examples.
[0042] Also, the PA may include multiple active devices. This is
called a multi-stage PA. The discussion will be in terms of one
active device, but it will be understood by those skilled in the
art that this discussion could be applied to multi-stage PA's.
[0043] Since the matching circuits and components are on a common
substrate 16, the impedance matches do not have to be to the
industry standard 50 ohms. Instead, the impedance match can be from
the natural output impedance, Z.sub.o, of one component to the
natural input impedance, Z.sub.i, of the next component.
[0044] For example, referring again to FIG. 1, if the PA unit 20
has an output impedance of about 2.5 ohms and the isolator unit 24
has an input impedance of about 12.5 ohms, the PA-to-isolator
matching circuit 41 will match the impedance from 2.5 ohms at the
PA unit 20 to 12.5 ohms at the isolator unit 24. This is in
contrast to the prior art. In the prior art, the PA unit would
typically have its own substrate, and the isolator unit would
typically have its own substrate. The PA unit would have its own
matching circuit which would match from the output of the PA (e.g.,
2.5 ohms) up to 50 ohms. The isolator unit would have its own
matching circuit which would match from 50 ohms down to the
isolator (e.g., 12.5 ohms). There would be additional loss in the
signal in this match up to 50 ohms from 2.5 ohms and back down to
12.5 ohms from 50 ohms.
[0045] A further advantage in matching from the natural ouput
impedance of one device to the natural input impedance of another
device is that a simpler topology in the matching network may often
be used when Zo and Zi are closer in value than they are to the
industry standard 50 ohms, for example. A simpler matching network
will result in less added variation due to component variation than
does a more complex network. In the limit where, for example,
Zo=Zi, no matching network is needed between adjacent devices in
the signal path. In the prior art, each device is typically matched
to the industry standard 50 ohms.
[0046] Referring again to FIG. 1, the duplexer 28 is a low loss
tunable duplexer, as described in U.S. patent application Ser. No.
09/912,753 filed on Jul. 24, 2001. Ferro-electric components such
as ferro-electric capacitors are used to tune the duplexer.
[0047] The integrated antenna interface unit has significantly less
loss in the transmit path than non-integrated transmit chains.
Integrating components, the PA, for example, eliminates lossy
attachments, which are described in U.S patent application Ser. No.
09/912,753 filed on Jul. 24, 2001. Specifically, an electrical
connection between the PA substrate and the common substrate is
eliminated. In the prior art, the PA is typically manufactured on
its own substrate. When a communication device is made,
incorporating the PA, an electrical connection must be made between
the PA substrate and the common substrate. Whether this is
accomplished by surface mount technology (SMT), hand soldering,
wire bonding, or some other attachment method, attachment losses
are added. By mounting the PA directly on the common substrate,
these losses are avoided.
[0048] Referring now to FIG. 2, a PA matching circuit 48 is shown.
A PA 50 has an input 52 and an output 54. In a preferred
embodiment, the output 54 is coupled to a first capacitor 56. The
first capacitor 56 is also coupled to ground. The output 54 is also
coupled to an inductive element 58. The inductive element 58 may be
a lumped element inductor, a microstrip line, or any other
inductive element known in the art. The inductive element 58 is
also coupled to a second capacitor 60. The junction between the
inductive element 58 and the second capacitor 60 forms the output
65 of the PA matching circuit 48. The output 54 of the PA 50 is
also coupled to a bias circuit. The bias circuit typically
comprises an inductor 68, a third capacitor 71 and a voltage source
74.
[0049] Another example matching circuit topology is shown in FIG.
3. The matching circuit 72 is similar to the matching circuit 48
shown in FIG. 2, except that the matching circuit 72 in FIG. 3 has
an additional inductive element 74 and an additional capacitor 76.
Also, the output 78 of this matching circuit 72 is at the junction
of the inductive element 74 and the capacitor 76. Any or all of the
inductive and capacitive components may be tunable.
[0050] It will be understood by those of skill in the art that
different matching circuit topologies might be used to implement
the PA matching circuit. In general a more complex matching circuit
will allow for greater control in the match at the expense of added
insertion loss (I.L.) due to finite component Q, as well as greater
cost and increased board space.
[0051] Referring again to FIG. 1, the PA is placed directly on the
substrate 16, and the matching circuit described with reference to
FIG. 2 is fabricated directly on the substrate 16. The capacitors
may be fabricated directly on the substrate 16 as interdigital
capacitors, gap capacitors or overlay capacitors, as is well known
in the art. By fabricating the PA unit 20, the PA-to-isolator
matching circuit 41 and the isolator unit 24 directly on the same
substrate 16, attachment losses are avoided, in addition to the
previously described losses resulting from matching impedances up
to and back down from 50 ohms. In the prior art, the separate
substrates for the PA unit and the isolator unit must be attached
electrically and mechanically to a common substrate or board. There
are losses associated with attachment of these additional
substrates. Finally, there is additional loss in the electrical
line connecting the separate substrates on the common substrate or
board. By combining the PA and isolator onto a common substrate
these losses are eliminated or significantly reduced.
[0052] Referring again to FIG. 2, the capacitors 56 and 60 may be
tunable, using low loss tunable ferro-electric materials and
methods as described in U.S. patent application Ser. No. 09/912,753
filed on Jul. 24, 2001, and Ser. No. 09/927,136 filed on Aug. 10,
2001, hereby incorporated by reference. This would reduce the loss
even further, by providing for an optimum impedance match. The
matching circuits shown in FIGS. 2 and 3 are used to match a single
band, such as the PCS band, or the cellular band. Presently, these
matching circuits can achieve a tunability of at least 15%.
[0053] This allows for tuning even over several international PCS
bands, such as from the India PCS band to the U.S. PCS band. To
tune over a wider frequency, for example, from the U.S. PCS band at
about 1900 MHz to the U.S. cellular band at about 800 MHz, the
PA-to-isolator matching circuit has to have more tunability.
[0054] For tuning a PA over more than one PCS band, the input
matching circuit may need tuning as well. Whether tuning the input
matching circuit is necessary or not can be determined on a case by
case basis. The same technique as used for the output matching
circuit is used in this case.
[0055] Increased tunability is attained by adding
micro-electro-mechanical switches (MEMS) to the matching circuit.
Referring now to FIG. 4A, a multiband PA matching circuit 31 is
shown. The matching circuit 31 is similar to that of FIG. 2, except
that several additional components have been added, with the
ability to switch those components in and out of the circuit 31
with MEMS. The output 35 of a PA 33 is coupled, as in FIG. 2, to a
first capacitor 37 and to a first inductive element 39. The first
inductive element 39 is coupled to a second capacitor 43. But here,
the output 35 of the PA 33 is also coupled to a first MEMS 45 for
selectively coupling to a third capacitor 47. The first inductive
element 39 and the second capacitor 43 are also coupled to a second
MEMS 80 for selectively coupling to a fourth capacitor 83. These
switches 45 and 80 and capacitors 47 and 83 change the capacitance
of the matching circuit 31.
[0056] Additionally, the first inductive element 39 is coupled at
either end to MEMS 86 and 89 for selectively coupling to a second
inductive element 92. These switches 86 and 89 and inductive
element 92 change the inductance of the matching circuit 31. In
this way, the matching circuit 31 can be used to match the PA 33
for use at either cellular or PCS bands. It will be understood that
the techniques and devices described here could be used to match at
other bands than the cellular and PCS bands. The cellular and PCS
bands are chosen as examples. It will also be understood that other
matching circuit topologies can be chosen.
[0057] Referring again to FIG. 4A, a multi band PA matching circuit
31 is shown which is similar to the single band PA matching circuit
described with reference to FIG. 2. As stated the multi band PA
matching circuit 93 has an advantage in that it is tunable over a
broader range of frequencies, due to the addition of MEMS switches
86, 89, 45 and 80 and accompanying components. Tunable capacitors
37 and 43 and tunable reactive element 39 can be used to fine tune
over a specific frequency band. The specific band is selected by
MEMS switches 86, 89, 45, and 80.
[0058] In addition to MEMS switches 86, 89, 45, and 80, the multi
band PA matching circuit 93 has additional capacitors 47 and 83 and
an additional reactive element 92. Capacitor 83 is connected in
series with capacitor 43 and in series with MEMS switch 80. When it
is desired to switch to another band, such as, for example, another
PCS band, MEMS switch 80 is activated, coupling capacitor 83 to
capacitor 43 and reactive element 39 for changing the impedance of
matching circuit 93. Similarly, MEMS switch 45 can be activated to
couple capacitor 47 to capacitor 37 and reactive element 39 for
changing the impedance of matching circuit 93. Also similarly, MEMS
switches 86 and 89 can be activated to couple reactive element 92
in parallel to reactive component 39 for changing the impedance of
matching circuit 93.
[0059] An alternative configuration of reactive components 92 and
39 and MEMS switches 86 and 89 is shown in FIG. 4B. In FIG. 4B MEMS
switches 86 and 89 are coupled to reactive elements 92 and 39 such
that only one of reactive elements 92 and 39 is coupled to
capacitors 37 and 43. Reactive element 39 can be switched out of
the circuit, so that it is disconnected at both ends, whereas, in
FIG. 4A, reactive element 39 is always coupled to the circuit at
capacitors 37 and 43. Reactive element 92 only is switched in and
out of the circuit. Note that in both FIGS. 4A and 4B any of the
elements 92, 39, 47, 37, 83 and 43 may be tunable. At least one is
tunable, but as few as one, or all of them, may be tunable.
[0060] For handset applications, the MEMS switches described here
should have the lowest practical loss, e.g., DC resistance less
than about 0.01 ohms. Switching speed is not critical so long as it
is less than about 1.0 ms. Clearly, other applications may require
other critical specifications on the MEMS switches.
[0061] Referring now to FIG. 5, an isolator matching circuit will
now be described. An input port 97 is coupled to a PA (not shown),
to a first impedance element 99 and to a second impedance element
101. The first and second impedance elements 99 and 101 form an
input matching circuit for the isolator 95. The second impedance
element 101 is coupled to ground, and the first impedance element
99 is couple to the isolator 95 for transmitting a signal from a PA
(not shown) to the isolator 95. Both the first and second impedance
elements may be ferro-electric tunable components, as described in
U.S. patent application Ser. No. 09/927,136 filed on Aug. 10,
2001.
[0062] An output of the isolator 95 is coupled to a third impedance
element 103, which is coupled to a fourth impedance element 105.
The third and fourth impedance elements 103 and 105 together form
an output matching circuit and an output port 107 for the isolator
95. The output port 107 is coupled to a duplexer (not shown). Both
the third and fourth impedance elements may be ferro-electric
tunable components, as described in U.S. patent application Ser.
No. 09/927,136 filed on Aug. 10, 2001.
[0063] An isolation port 104, is coupled to an impedance element
109. The impedance element 109 is coupled to another impedance
element 115 and to a resistor 118. Together the impedance elements
109 and 115 and the resistor 118 comprise an isolation matching
circuit.
[0064] It will be understood by one of skill in the art that the
input, output and isolation matching circuits described with
reference to FIG. 5 using "L" matching sections are illustrative
only. Other topologies for these matching circuits could be used,
such as, for example, parellel LC circuits, "T", or Pi networks, as
described in U.S. patent application Ser. No. 09/927,136 filed on
Aug. 10, 2001.
[0065] Advantageously, each of the impedance elements 99, 101, 103,
105, 109, and 115 are preferably formed directly on the common
substrate described with reference to FIG. 1. This reduces losses
associated with attaching separate units to the substrate, reduces
cost and eliminates the need to match components up to the 50 ohm
industry standard.
[0066] Regarding the PA, its characteristic output impedance for
CDMA handsets is typically about 2-4 ohms near the maximum output
power level required of it. The isolator characteristic impedance
is typically about 8-12 ohms. Filters can be designed with input
and output impedances that can take on a broad range of values.
Since duplexers and diplexers are made primarily of filters, they
can be designed to allow for a broad range of input and output
impedances. Thus, they can be designed to match to whatever
impedance is convenient based on the rest of the circuit.
[0067] Referring now to FIG. 6A, a preferred LNA matching circuit
117 will now be described. An input port 118 is coupled to a first
inductor 121 and to a capacitor 124. The capacitor is coupled to a
second inductor 127. The second inductor 127 is coupled to a third
inductor 130 and to an LNA 133.
[0068] The matching circuits will be used to match the impedance
between the various parts to avoid or reduce power loss in the
signal travelling from one part to the other. For LNA applications,
there is another purpose. For LNA applications, impedance
transforming networks or circuits are used primarily to maintain an
optimum noise impedance match between the input signal source and
the active device chosen for the LNA. In fix-tuned circuits, the
optimum noise impedance match is obtained at one frequency and is
dependent on both temperature and component variations. In the
tunable circuit approach described here, the optimum noise
impedance match can be made adjustable to cover multiple bands or a
wider frequency range than is possible in the fix-tuned case. An
added advantage in using tunable components is the ability to
compensate for temperature variations.
[0069] The introduction of f-e or other tunable components allows
for increased flexability in the design of LNA's. In the
conventional design using fixed elements, one must usually
trade-off optimum noise figure and maximum gain. With tunable
components, one can allow for cases where the input matching
circuit can be varied from the minimum noise figure and the maximum
gain, as desired.
[0070] A tunable optimum noise figure will now be described with
reference to FIG. 6B. FIG. 6B is a graph showing noise figure 120
plotted against frequency 122. Typically, such as, for example, in
a CDMA wireless communication device, there will be a maximum noise
figure 126 specified for a given design of an LNA. The maximum
noise figure specified is shown as a horizontal dashed line 126. A
curve showing a typical noise figure response 128 is shown as the
solid curve.
[0071] Typically, the LNA and its matching circuits will be
designed so that the noise figure response 128 will be below the
maximum noise figure 126 at an operating frequency, f.sub.o 130. A
tunable LNA matching circuit allows the LNA noise figure response
128 to be tuned over frequency. The tuned noise figure response 132
and 134 is represented by two dashed curves of a similar shape to
that of the typical noise figure response 128. By tuning the noise
figure response at 132 and 134, the noise figure response can be
made to be below the maximum noise figure 126 at alternate
operating frequencies f.sub.1 138 and f.sub.2 140. It will be
understood that f.sub.1 138 and f.sub.2 140 are chosen as
representative frequencies only. The noise figure response can be
tuned over a broad range of frequencies. Additionally, it will be
understood by one of skill in the art that MEMS switches can be
added to the LNA matching circuit to further broaden the range of
tunability of the noise figure response.
[0072] Referring now to FIG. 7, a preferred antenna matching
circuit will now be described based on a CDMA handset. An antenna
136 is coupled to a first inductor 139 and to a second inductor
142. The first inductor 139 preferably has an inductance equal to
about 8.2 nH. The second inductor 142 preferably has an inductance
equal to about 3.9 nH.
[0073] The second inductor 142 is coupled to a first capacitor 145
and a second capacitor 148. The first capacitor 145 preferably has
a capacitance equal to about 0.5 pF. The second capacitor 148
preferably has a capacitance equal to about 2.7 pF. It will be
understood that other component values and matching circuit
topologies can be used.
[0074] One side of the second capacitor forms an input and output
port 149 for the antenna matching circuit for coupling to a
duplexer (not shown), diplexer (not shown), multiplexer (not shown)
or other type of filter (not shown).
[0075] The antenna matching circuit will typically be a pi or T
circuit with an L-C ladder making it a higher order match. This
gives more tolerance for impedance variation. Typically, the
antenna in a system will be matched to 50 ohms. There may be,
however, an ideal impedance for a given antenna that is other than
50 ohms, though 50 ohms is common for test devices.
[0076] For example, a commonly used antenna for wireless
communication devices may have an input impedance of 30 ohms. As
previously mentioned, the PA may have an ouput impedance of about 2
ohms. The isolator may have an ouput impedance of about 12.5 ohms.
The diplexer and duplexer filters can easily accommodate a wide
range of impedances.
[0077] So the PA-to-isolator match is from about 2 ohms at the PA
to about 12.5 ohms at the isolator. The isolator-to-duplexer match
is from about 12.5 ohms to about 12.5 ohms. The duplexer is at
about 12.5 ohms. So the duplexer-to-diplexer match is about 12.5 to
about 12.5 ohms. The diplexer and duplexer inputs and outputs are
at about the same impedance, for example, about 12.5 ohms. The
diplexer-to-antenna matching circuit may be a match from about 12.5
ohms at the diplexer to about 30 ohms at the antenna. Each of these
matching circuits, plus the diplexer and the duplexer may be f-e
tunable.
[0078] At mentioned above with reference to FIG. 1, it will be
understood that a common substrate may include many different
combinations of the parts mentioned above. In one embodiment, as
shown in FIG. 8, a common substrate 152 includes a duplexer 154, an
isolator 156, a PA 157 and the requisite matching circuits (not
shown). In another embodiment, as shown in FIG. 9, a common
substrate 160 includes an antenna matching circuit 163, a diplexer
166 and a duplexer 169. In yet another embodiment, as shown in FIG.
10, a common substrate 172 includes an antenna matching circuit
175, a diplexer 178 and two duplexers 181 and 184. In still another
embodiment, as shown in FIG. 11, a common substrate 186 includes an
antenna matching circuit 188, a diplexer 190, two duplexers 192 and
194, two isolators 194 and 196, two PA's 198 and 200 and two LNA's
202 and 204. In another embodiment, as shown in FIG. 12 a common
substrate 206 includes everything mentioned above with reference to
FIG. 11, except the antenna matching circuit 188. In another
embodiment, as shown in FIG. 13 a common substrate includes
everything mentioned above with reference to FIG. 10, except the
antenna matching circuit 175.
[0079] The integration of a PA module, isolator and duplexer for a
CDMA TX chain removes the requirement that each stand-alone device
be matched at 50 ohms at the input and output. By allowing for a
more gradual impedance match (from about 2 ohms to about 30 ohms in
the example given) one can reduce match-induced losses.
Additionally, the f-e tunable components are exposed to a lower rf
voltage, for a given power.
[0080] The reduced rf voltage, for a given power, reduces
non-linear distortion, because f-e films are typically non-linear.
Alternatively, a f-e component can be subjected to increased power
while maintaining an acceptable level of non-linear distortion.
Thus, designing integrated components that operate at lower input
and output impedances allows for f-e components to be incorporated
in applications where higher power levels are required than
possible with f-e components matched to the industry standard 50
ohms.
[0081] Fabrication on a common substrate further reduces losses
that naturally arise when the components involved are packaged and
mounted individually on a printed wire board (pwb).
[0082] By reducing Tx chain losses the Tx chain specifications can
more easily be satisfied. This means that the specification for one
or more of the parts involved can be relaxed. For example, the PA
or other high value part specifications can be relaxed. A high
value part is a part with one or more of the following
characteristics: high cost, high performance, high level of
difficulty in meeting specifications such as gain, power out,
stability, ACPR, over temperature, and unit-to-unit
repeatability.
[0083] Since the specifications on the PA, for example, can be
relaxed, there are many possible benefits. For example, the PA may
be able to meet specifications while consuming less power. This
results in longer talk times or longer standby times or both. In
another example, since Tx chain losses are reduced, a wireless
handset manufacturer may be able to meet specifications with a PA
that has less stringent tolerances or requirements. The handset
manufacturer may be able to choose a cheaper PA, reducing the cost
of wireless handsets. These benefits of reduced Tx chain losses are
given as examples only. It will be understood by those skilled in
the art that other benefits will arise from reduced Tx chain
losses. It will further be understood that these benefits can be
utilized to improve wireless communication devices in ways other
than those mentioned here.
* * * * *